sign in sign up
<<
Home , Normative, Set (psychology), Social norm

Chicago-Kent Review

Volume 74 Issue 4 Symposium on Law, , and Article 6 the

October 1999

The Limits of Social Norms

Jeffrey J. Rachlinski

Follow this and additional works at: https://scholarship.kentlaw.iit.edu/cklawreview

Part of the Law Commons

Recommended Citation Jeffrey J. Rachlinski, The Limits of Social Norms, 74 Chi.-Kent L. Rev. 1537 (2000). Available at: https://scholarship.kentlaw.iit.edu/cklawreview/vol74/iss4/6

This Article is brought to you for free and open access by Scholarly Commons @ IIT Chicago-Kent College of Law. It has been accepted for inclusion in Chicago-Kent Law Review by an authorized editor of Scholarly Commons @ IIT Chicago-Kent College of Law. For more , please contact [email protected], [email protected]. THE LIMITS OF SOCIAL NORMS

JEFFREY J. RACHLINSKI*

No can function if it cannot constrain the self-serving behavior of its members. that cannot control socially destructive behavior collapse into dysfunction; they become dangerously -ridden, as in some of America's inner cities, or completely anarchic, as in parts of the Balkans and central Africa.' Clear rules enforced by legal sanctions deter a great deal of socially destructive conduct, but social norms enforced by informal sanctions might create even more powerful constraints. 2 If so, then gaining control over dysfunctional societies might depend more upon using or manipulating social norms than upon enforcing the law. Decades of research conducted by social psychologists on social norms, however, suggests three important obstacles to the use of social norms: First, antisocial norms, once established, are hard to dislodge; second, even if people adhere to positive social norms, determining when they are triggered is difficult; third, subtle aspects of situations can induce antisocial conduct, seemingly even against social norms. Most legal scholarship addressing the control of antisocial behavior addresses formal sanctions rather than social norms.3 This is not surprising; formal sanctions are more familiar to lawyers and punishing undesirable antisocial conduct is the most straightforward means of maintaining . Formal sanctions almost certainly provide a critical degree of control over antisocial conduct, but they are costly, both to society and to the individual punished. 4 Incarceration, in particular, can produce unintended costs to

* Professor of Law, Cornell Law School. Ph.D., Psychology, Stanford University, 1994; J.D., Stanford University, 1993; B.A., The Johns Hopkins University, 1988; M.A., Psychology, The Johns Hopkins University, 1988. I thank Chris Guthrie and Emily Sherwin for their helpful comments on this paper. 1. See Tracey L. Meares & Dan M. Kahan, Law and (Norms of) Order in the Inner City, 32 L. & Soc'Y. REv. 805, 812-13 (1998). 2. See generally Dan M. Kahan, Social Meaning and the Economic Analysis of Crime, 27 J. LEGAL STUD. 609 (1998). 3. See Robert C. Ellickson, Law and Discovers Social Norms, 27 J. LEGAL STUD. 537, 542-43 (1998). 4. See John J. Donohue III & Peter Siegelman, Allocating Resources Among Prisonsand Social Programs in the Battle Against Crime, 27 J. LEGAL STUD. 1,3-6 (1998). CHICAGO-KENT LAW REVIEW [Vol. 74:1537

individuals, especially in in which the rate of incarceration is high.5 Partly in response to the staggering costs of incarceration, some legal scholars argue that informal social norms provide a better means of controlling antisocial conduct. 6 This builds upon a growing body of legal scholarship concerning social norms.7 This new "law and social norms" scholarship adopts a more subtle approach to than identifying situations that should trigger formal rewards and sanctions. It accounts for the prominence of group membership and social approval as sources of rewards and . 8 This new field has developed several positive and implications of social norms. On the positive side, the scholarship asserts that groups develop and enforce norms of conduct apart from formal legal sanctions.9 Perhaps more significantly, this scholarship might hold valuable normative lessons for policymakers interested in reducing crime as an alternative to incarceration. The law and social norms scholarship identifies two ways in which reformers can take advantage of the power of social norms. First, changes in law can influence social norms. For example, a law against smoking in public places had a dramatic effect on smokers, not because of the formal penalty for public smoking (which is hardly ever imposed) but because it empowered nonsmokers to levy social sanctions on smokers.1 Changes in law might also inspire changes in preferences. 1 Scholars refer to this as the "expressive"

5. See Meares & Kahan, supra note 1, at 813. 6. See generally Dan M. Kahan, , Social Meaning, and Deterrence, 83 VA. L. REV. 349 (1997); Meares & Kahan, supra note 1, at 812. 7. See, e.g., Lawrence Lessig, The of Social Meaning, 62 U. CHI. L. REV. 943 (1995); Richard H. McAdams, The Origin,Development, and Regulation of Norms, 96 MICH. L. REV. 338 (1997); Richard A. Posner, Social Norms and the Law: An Economic Approach, 87 AM. ECON. REV. 365 (1997); Eric A. Posner, The Regulation of Groups: The Influence of Legal and Non-Legal Sanctions on Action, 63 U. CHI. L. REV. 133 (1996); Cass R. Sunstein, Social Norms and Social , 96 COLUM. L. REV. 903 (1996). Owing to the number of scholars at the University of Chicago doing this work, some have referred to it as the "New Chicago School." See, e.g., Lawrence Lessig, The New Chicago School, 27 J. LEGAL STUD. 661 (1998). 8. See McAdams, supra note 7, at 341-42. 9. See generally ROBERT C. ELLICKSON, ORDER WiTHouT LAW: How NEIGHBORS SETrLE DISPuTEs (1991); Lisa Bernstein, Opting Out of the Legal System: Extralegal ContractualRelations in the Diamond Industry, 21 J. LEGAL STUD. 115 (1992); Eric A. Posner, supra note 7, at 137-44. 10. See Robert Cooter, Expressive , 27 J. LEGAL STUD. 585, 594-95 (1998). 11. See Cass R. Sunstein, Endogenous Preferences, EnvironmentalLaw, 22 J. LEGAL STUD. 217, 230-34 (1993). 2000] THE LIMITS OF SOCIAL NORMS function of law.12 Second, subtle and inexpensive changes in a social setting can signal different social norms. For example, fixing broken windows, cleaning up trash, erasing , and planting flowers in a neighborhood can have a dramatic effect on crime rates because tidiness can "signal" a 's intolerance for lawbreaking. 3 The interaction between law and social norms arguably identifies ways to reduce crime that are less expensive and more humane than traditional approaches. A change in law that does not need to be enforced, or is only minimally enforced, is not costly to society or to the individuals who are deterred from breaking the law by a new social . Likewise, cleaning neighborhoods is much more humane than identifying and sentencing lawbreakers to lengthy periods of incarceration. Historically, those who have argued against incarceration have advocated social programs designed to fight poverty as a substitute crime control measure. 14 Such programs are expensive, and do not always succeed, or generate unwanted consequences. The law and social norms scholarship suggests novel alternatives to the choice between expensive social programs and incarceration. Manipulating social norms to reduce antisocial conduct will only be a successful strategy if the underlying describing how social norms behavior are both accurate and reasonably complete. Although the law and social norms scholarship relies largely on observational and anecdotal support, it presents a case that social norms influence behavior in ways that policymakers can use. The diamond industry is probably governed by the informal norms Professor Bernstein identifies, 5 sprucing up a community probably reduces crime as Professor Kahan argues,'16 and changes in the law probably affects people's of conduct such as dueling and wearing hockey helmets. 7 If the theories explaining these phenomena are incomplete, however, then the law and social norms scholarship is merely a post hoc effort to accommodate some anomalous phenomena. 8 The law and social norms theories might

12. See Cooter, supra note 10, at 595. See generally Cass R. Sunstein, On the Expressive Function of Law, 144 U. PA. L. REV. 2021 (1996). 13. See Kahan, supra note 6, at 352-61. 14. See, e.g., Donohue & Siegelman, supra note 4. 15. See Bernstein, supra note 9. 16. See Kahan, supra note 6, at 352-61. 17. See Lessig, supra note 7, at 968-71. 18. See Sunstein, supra note 7, at 945. CHICAGO-KENT LAW REVIEW [Vol. 74:1537 not generalize to new situations, making them useless to policy- makers. Law and social norms so far has relied upon rational choice theory and theory for its foundation,19 but sociologists and social psychologists have a long of studying social norms. even includes a school of known as "," which is stunningly similar to the signaling theories in the new legal scholarship on social norms.20 For their part, social psychologists have been studying social norms and social influence since the 1930s.21 Decades of social psychological research on social norms merits more than just a tip-of-the-hat from law and social norms scholars. 22 As early as 1936, a prominent social psychologist, Muzafir Sherif, published a monograph describing empirical research on the origins and implications of social norms.23 One of the founders of modern , Kurt Lewin, demonstrated how to manipulate social settings to alter social norms during World War II as part of an effort to reduce demand for scarce products needed for the war effort.2 4 Social psychologists conducted decades of research on social norms and social influence that followed these early leaders, which could enrich the law and social norms scholar's anecdotal with rigorous . The social psychological research, in , provides some systematic support for the theories proposed by the social norms scholars. Social psychologists have conducted numerous empirical studies which support two basic tenets of the law and social norms scholars: (1) that groups develop and impose norms on their members, and (2) that the apparent behavior of others can alter the social meaning of a situation in ways that profoundly affect behavior. This research provides direct support for the signaling theories

19. See, e.g., Randal C. Picker, Simple in a Complex World: A Generative Approach to the Adoption of Norms, 64 U. CHI. L. REV. 1225 (1997). 20. See generally JOEL M. CHARON, SYMBOLIC INTERAcTIONISM: AN INTRODUCTION, AN INTERPRETATION, AN INTEGRATION (6th ed. 1998); , THE PRESENTATION OF SELF IN EVERYDAY LIFE (1959) (providing an early account of the importance of symbolic actions). 21. See, e.g., MUZAFIR SHERIF, THE PSYCHOLOGY OF SOCIAL NORMS (1936). 22. Not being a sociologist, I do not purport to speak for the of the sociological work; I limit my discussion to social psychology. 23. See SHERIF, supra note 21. 24. See Kurt Lewin, Group Decision and , in READINGS IN SOCIAL PSYCHOLOGY 197, 197 (Eleanor E. Maccoby et al. eds., 3d ed. 1958). 20001 THE LIMITS OF SOCIAL NORMS developed by the law and social norms scholars. 5 Social psychology, however, suggests several caveats to the law and social norms scholars as well. First, people are much more active and dynamic interpreters of their surroundings than the law and social norms theories have argued. One consequence of this dynamism is that predicting how people interpret social situations is difficult; people do not simply conform to the of a group majority in all contexts. Second, subtle aspects of social situations can induce people to behave in ways that are apparently contrary to the social norms they espouse. Antisocial conduct is frequently not the product of norms, but of a situation that channels people to commit these acts. Third, social norms are always in tension with each other, which makes predictions as to which norm will dictate social behavior invariably unstable. These caveats suggest that the law and social norms scholarship needs to develop a richer theory of social influence than rational- choice models and will provide. The caveats of social psychology also raise questions about one of the normative implications of the new scholarship. If antisocial behavior is as deeply rooted in social situations as the social psychological research suggests, then tidying up a neighborhood might be doing more to move crime around than reducing its overall prevalence. Although historic "great society" programs to attack the roots of crime have had only mixed success and have developed a bad name, they may have had the right goal of altering the basic social situations that inspire crime in the first place. Furthermore, if social behavior is as difficult to predict as the social psychological research suggests, then crime-control programs based on signaling theories might have only marginal success in the long run. It should not be surprising that preferences for prosocial and antisocial conduct are as difficult to manage and predict as other preferences. If government cannot be trusted to effectively manage the supply of hammers and zinc (and hence markets are superior), how can it be trusted to manage the supply of social norms? Social psychology suggests that all such preferences and the behaviors that accompany them are phenomenally hard to regulate. This Essay documents the caveats that social psychological

25. The theory that law has an expressive function has not really been tested by social psychologists. Scholars in the law and society or fields have more direct on this point and hence they are beyond the scope of this Essay. CHICAGO-KENT LAW REVIEW [Vol. 74:1537 research suggests for the value of social norms for policymakers. Part I describes the newer research on social norms. Part II reviews and interprets the original research by social psychologists on social norms. Part III identifies the consistencies and caveats between the law and social norms scholarship and the social psychological scholarship with specific reference to the merits of crime-control measures.

I. SOCIAL NORMS AND RATIONAL CHOICE Societies frequently manipulate the rate of pro-social and antisocial behavior by adopting formal schedules of rewards and punishments. People also conform to norms of behavior, even when the costs and benefits of doing so are not readily apparent. Consequently, societies can affect the rate of pro-social or antisocial conduct by identifying and manipulating the operation of these norms. In fact, in many instances, societies can have a greater influence on conduct by altering social norms than they can by altering the formal schedule of costs and benefits.

A. Rational Choice and Deterrence

It is undeniably true that society can influence behavior by altering formal costs and benefits. Punishing traffic violations with a death sentence would doubtless deter speeding more than would a fifty dollar fine. Likewise, providing a ten dollar rebate per aluminum can brought to a recycling center would vastly reduce the number of discarded cans relative to a five cent rebate. Formal incentives clearly provide societies with a powerful means of controlling social behavior. In modern society, positive incentives are usually financial, and penalties consist primarily of a mix of fines and incarceration. Relying on formal subsidies and penalties creates several unwanted consequences, however. First, a penalty or subsidy might be excessive, deterring or promoting conduct that is otherwise desirable.26 For example, imposing a massive fine for double-parking could deter people from parking in such a fashion in emergencies.27

26. See Gary S. Becker, Crime and : An Economic Approach, 76 J. POL. ECON. 169 (1968). 27. See A. MITCHELL POLINSKY, AN INTRODUCTION TO LAW AND ECONOMICS 76-78 (2d ed. 1989). 20001 THE LIMITS OF SOCIAL NORMS

Second, penalties can create perverse incentives. For example, punishing people for killing endangered species on their property creates incentives for landowners to ensure that no endangered species will inhabit their property in the first place.28 Third, penalties and subsidies are themselves expensive. For example, studies of the social costs and benefits of imprisonment indicate that the social gains 29 from incarcerating criminals can be fleeting.

B. Social Norms The problems with using formal incentives to regulate behavior suggest other mechanisms to control antisocial conduct might be more effective. The law and social norms scholarship provides many examples of situations in which people behave in ways that seem, at least superficially, inconsistent with their own self-interest. 30 Even in sterile settings designed by psychologists and behavioral economists to be devoid of social context, people allocate rewards between themselves and others according to equitable norms, even at their own expense.3' Social norms operate somewhat independently from formal law. commonly track social norms; murder, burglary, and many are both illegal and socially inappropriate.32 Sometimes, however, social norms and law conflict. In such instances, norms frequently influence behavior more than the law.33 For example, Professor Ellickson has shown that community norms among neighbors on allocating costs of stray cattle in Shasta County, California, operate independently of the law's allocation of these costs.34 An inconsistency between law and social norms makes enforcing formal laws against antisocial conduct difficult.35 Absent some power to manipulate social norms, these

28. See generally Richard A. Epstein, Babbitt v. Sweet Home Chapters of Oregon: The Law and Economics of HabitatPreservation, 5 SUP. CF. ECON. REV. 1 (1997); Stephen Polasky et al., Endangered Species Conservation on Private Land, 15 CONTEMP. ECON. POL'Y 66 (1997); Richard L. Stroup, The Economics of Compensating Property Owners, 15 CoNTEMP. ECON. POL'Y 55 (1997). 29. See Donohue & Siegelman, supra note 4, at 30-40. 30. See McAdams, supra note 7, at 340-41; Sunstein, supra note 10, at 904-14. 31. See Colin Camerer & Richard H. Thaler, Anomalies: Ultimatums, Dictators, and Manners, 9 J. ECON. PERSP. 209, 216-18 (1995). 32. See McAdams, supra note 7, at 347-48. 33. See id. at 348. 34. ELLICKSON, supra note 12. 35. See McAdams, supra note 7, at 348. CHICAGO-KENT LAW REVIEW [Vol. 74:1537 have only limited value for reformers. Identification of divergences between law and norms might indicate how best to allocate an enforcement budget, but this is a relatively weak implication. Scholars have argued, however, that the law interacts with social norms to influence behavior in two productive ways: First, laws can direct or strengthen existing social norms;36 and second, 37 social policies can change the social meaning of a behavior. Law has an expressive function that can motivate a change in social norms.3 8 Attaching a criminal penalty to conduct can inspire social condemnation. Criminalizing undesirable conduct to support a can embolden people to levy informal sanctions against a violator and signal potential violators that their conduct will draw a severe social sanction. Ordinary citizens might begin to feel entitled to be free from the costs of the undesirable conduct that violators inflict on them.39 Consequently, even in the absence of enforcement, the mere act of criminalizing conduct can reduce its prevalence. For example, as Robert Cooter has observed, even though laws against smoking in public are almost never enforced, is widespread. 4° Cooter suggests that labeling the behavior as a crime will heighten potential violators' fear of social sanction, and might 41 also encourage a real increase in social sanctions against violators. Along a similar line, laws and policies can take advantage of social norms to change the meaning of social behaviors in ways that discourage antisocial conduct. For example, sprucing up a rundown community by planting flowers, fixing broken windows, and cleaning up graffiti can change the signal that the community sends to potential lawbreakers. 42 A neat, well-tended neighborhood conveys the impression that its residents are willing to take other actions to maintain a high of life, such as cooperating with police to apprehend lawbreakers. Neighborhood improvement projects need not actually increase a community's anticrime efforts, but it signals a greater likelihood that the community will punish lawbreakers.

36. See generally Cooter, supra note 10. 37. See generally Kahan, supra note 9. 38. See generally Cooter, supra note 10. 39. See Sunstein, supra note 14, at 230-34. 40. See Cooter, supra note 10, at 594-95. 41. Id. 42. See Robert C. Ellickson, Controlling Chronic Misconduct in City Spaces: Of Panhandlers,Skid Rows, and Public-Space Zoning, 105 YALE L.J. 1165, 1177-78 (1996); Kahan, supra note 9, at 369-71; James Q. Wilson & George Kelling, Broken Windows, ATLANTIC, March 1982, at 29, 29-30. 2000] THE LIMITS OF SOCIAL NORMS

Using the law to influence social norms raises the prospect of having a big effect at little cost. Creating a law against smoking in public and not enforcing it is practically a costless means of eliminating the problem of exposure to second-hand smoke. Reforms directed at changing social norms might also prove much more effective and have fewer unwanted side effects than enforcing formal penalties.

C. Why Do Social Norms Affect Behavior? Developing successful social policies that rely on social norms to reduce antisocial conduct requires an of how social norms influence behavior. Without this understanding, the of social norms risks becoming a of post hoc observations about behavior that lacks any predictive power.43 Merely attributing behavior that cannot easily be explained in terms of its short-term costs and benefits to some underlying norm is not helpful. The real value of social norms for legal scholarship lies in identifying social reforms that can be identified ex ante as having the ability to reduce antisocial conduct. Legal scholars have produced several different theories to explain the influence of social norms. First, group membership is often so valuable that the threat of expulsion might have a big effect on behavior. In the case of customary practices in industry groups, for example, compliance with the industry practices often requires that members forego strategic advantages that they might otherwise gain over others in exchange for continued membership in the group. 4 Because continued access to the benefits of group membership vastly outweighs the short-term benefits of violating the group's customs, the group members tend to conform. This account of social norms is really an extension of the that incentives . It simply incorporates the rewards and sanctions that accompany membership in a cooperative group into the analysis. As a related theory, some scholars have suggested that by others is such an important source of self-esteem for most people that groups of people have an inherent ability to influence behavior. 45 are social animals, and acceptance by others is itself important, whether or not this acceptance leads

43. See McAdams, supra note 7, at 341; Sunstein, supra note 10, at 945. 44. See Bernstein, supra note 9, at 134-35. 45. See McAdams, supra note 7. CHICAGO-KENT LAW REVIEW [Vol. 74:1537 directly to other rewards. Individuals might therefore be willing to engage in conduct that benefits a group at their own expense, purely to attain or to maintain the social contact and acceptance that accompanies group membership. This theory is also perfectly consistent with the incentives theory, but it recognizes social approval as an intrinsically important reward. Legal scholars have also recognized that group membership provides significant information to others.46 Because small, seemingly insignificant behaviors can signal membership in a group, people might engage in such behaviors purely to signal others that they belong to a certain group, or subscribe to a certain set of beliefs.47 For example, voting is often identified as a behavior that is difficult to explain in terms of costs and benefits, but can be explained in terms of signaling.48 By itself, voting is unlikely to be of much value to the voter. The act of voting, however, signals to neighbors and friends that one is a public-spirited member of the community. People conform their behavior to the standards of a group to signal their membership in that group and thereby achieve any benefits that such membership conveys. The theory that cleaning up a neighborhood deters crime takes advantage of these . Cleaning up trash and fixing broken windows is an effort to signal to potential lawbreakers that the community does not tolerate crime. This theory is obviously attractive to policymakers seeking to reduce antisocial conduct. The theory suggests cheap and effective mechanisms to reduce antisocial conduct.49 Although hiring someone to collect trash and erase graffiti is not free, it is apt to be much less expensive than either widespread incarceration or massive antipoverty efforts. The law and social norms scholarship thus presents an intriguing new direction for crime control. Neither incarceration nor antipoverty programs addresses social norms, and hence miss an opportunity to implement low-cost mechanisms to reduce crime.

46. See, e.g., Eric Posner, supra note 10. 47. See id. 48. See generally Richard L. Hasen, Voting Without Law?, 144 U. PA. L. REv. 2135 (1996). But see George A. Quattrone & Amos Tversky, Causal Versus Diagnostic Contingencies: On Self- and the Voter's , 46 J. PERSONALITY & SOC. PSYCHOL. 237, 244-47 (1984) (providing an of voting based on cognitive ). 49. See Kahan supra note 9; Meares & Kahan, supra note 1. 2000] THE LIMITS OF SOCIAL NORMS

II. SOCIAL INFLUENCE, SOCIAL MEANING, AND SOCIAL PSYCHOLOGY Social psychologists have long been interested both in social policies aimed at reducing crime and in social norms. Like the law and social norms scholars, they have criticized incarceration as ineffective and brutal and have criticized elaborate antipoverty programs. Yet, social psychologists have not generally endorsed social norms as a primary means of crime control. The may be that social psychological research identifies both the power and limits of social norms. Norms dominate human social behavior, but social psychological research demonstrates that their influence can be ephemeral and hard to predict. In short, research both supports the law and social norms scholarship and counsels caution in the application of its teachings. A review of the social psychological research reveals much support for the law and social norms scholarship. It also provides the basis for social psychologist's caution.

A. Social PsychologicalSupport for Social Norms Theory Groups exert a powerful influence on individual behavior. Groups of people spontaneously form social norms, with which their members conform. Group norms can also affect people's and beliefs. Even without a shared norm, the behavior of similar others has a profound effect on what people do and how they think because it provides an important source of information on how to interpret social situations. These observations support the law and social norms signaling theories. These observations also have support from social psychological research, beginning with the field's earliest experiments. 1. The Influence of Group Norms on Perception: Sherif's Autokinetic Effect The influence of others can influence how people see the world, even to the point of affecting basic human perception. Nearly seventy years ago, psychologist Muzafir Sherif argued that people naturally adhere to norms, even when judging mundane perceptual stimuli.5 0 To demonstrate this, Sherif documented what he referred to as the

50. SHERIF, supra note 21, at 89-112. CHICAGO-KENT LAW REVIEW [Vol. 74:1537

"autokinetic effect." In his study, Sherif asked people to stare at a point of light in an otherwise completely darkened room. Sherif informed his subjects that under these circumstances, most people perceive that the light has moved, even though it remains stationary. Sherif asked the subjects to identify how far the light seemed to move during a two-second interval, and then repeated the exposure several . Subjects viewed the stimuli either first alone and then in groups of two or three, or in groups of two or three first and then alone. In all cases, the subjects announced their estimates verbally. Sherif hypothesized that on their own, people would develop internal norms to guide them in deciding how far the light moved, and that their answers would remain within this internally set range. Sherif also reasoned that a group would create its own norm, to which individuals would then conform. Social norms clearly developed in Sherif's study. When responding in a group, subjects' responses coalesced into a much narrower range than when subjects responded alone. Subjects who responded first in groups quickly adapted to the group norm, and generally retained this norm when they began responding alone. Subjects who first responded alone also adjusted their own responses to conform to an emerging group norm. The results of Sherif's experiment show that even in a minimally social setting, groups create a social norm, the influence of which persists even in the group's absence. Psychologists have long used the power of group norms to induce pro-social behavior. For example, during World War II, psychologist Kurt Lewin used social influence techniques to encourage people to eat more organ meats, so as to reduce demand for meat.5 Lewin organized groups of women who were responsible for meal decisions in their homes to meet together and discuss the importance of in wartime. When Lewin induced each member of these groups to make a public commitment to using organ meats in at least one meal per week, he found much greater compliance with the request than when he did not insist on a public commitment. Public commitment to this social norm overcame resistance to the underlying request. The studies conducted by Lewin and Sherif revealed how powerful group norms can be. Sherif showed both that norms of behavior develop quickly and easily, and that these norms affect

51. Lewin, supra note 24. 20001 THE LIMITS OF SOCIAL NORMS behavior even after the group would seem to have no influence. Groups have an unspoken ability to control their members. Lewin's work showed that inducing public commitment to adhere to a group's norms increases the power of these norms even further. These studies support the idea that group norms can provide effective control over antisocial conduct. In fact, numerous self-help and self-improvement groups have taken advantage of the power of group influences to help people control their behavior. 52 Studies like those of Lewin and Sherif demonstrate that social norms merit the kind of that the new law and social norms scholars believe that they do. 2. The Power of Social Models: Social Cognitive Modeling In fact, social psychological research suggests that the law and social norms scholars do not go far enough. People's behavior has a profound influence on those around them, even when it does not convey information about a social norm. The tenet that observing another person, especially someone similar, engaging in a behavior makes someone more likely to engage in the same behavior is practically a law of social psychology.53 As one example, cited by law and social norms scholar Professor Kahan as empirical support for the power of social influence, 54 psychologist Philip Zimbardo induced people to destroy an automobile by providing a model who began to smash it with a sledgehammer.55 Passersby destroyed a similar automobile placed in a neighborhood in New York City even without a similar model, arguably because people in this neighborhood had (unfortunately) witnessed many examples of such activities.5 6 As Professor Kahan noted, Professor Zimbardo's study is impressionistic, rather than scientific.57 But it is just the tip of an iceberg of studies on the power of social models. Psychologist Albert Bandura conducted numerous systematic studies of the power of

52. See LEE Ross & RICHARD E. NISBETr, THE PERSON AND THE SITUATION: PERSPECIVES OF SOCIAL PSYCHOLOGY 222 (1991). 53. See ALBERT BANDURA, SOCIAL FOUNDATIONS OF THOUGHT AND ACTION: A SOCIAL COGNITIVE THEORY 47-48 (1986). 54. Kahan, supra note 6, at 355-56. 55. Phillip G. Zimbardo, The Human Choice: Individuation, Reason, and Order Versus ,Impulse, and Chaos, in NEBRASKA SYMPOSIUM ON 1969, at 237, 287-93 (William J. Arnold & David Levie eds., 1969). 56. See id. 57. Kahan, supra note 6, at 355-56. CHICAGO-KENT LAW REVIEW [Vol. 74:1537 social modeling on the level of aggression in children. 8 For instance, he demonstrated that exposing children to examples of aggressive behavior increased the level of aggression they exhibited. 9 In one of his studies, children punched a "bobo doll" (an inflatable clown weighted heavily on the bottom so that it always pops back up when knocked around) more often and more aggressively after watching a video of another engaging in a violent exchange with the doll than after watching a child engaging in more passive . 6° According to Bandura, social modeling influences behavior in two ways, both of which are basically consistent with the theories advanced by the social norms scholars in law. First, watching someone else engage in a certain behavior suggests that the behavior is acceptable under the circumstances. 61 The fact that some other child has endlessly pummeled a bobo doll without adverse consequences suggests to the child watching that he or she can also do so without penalty. This account supports the law and social norms theorists; watching someone else engage in a behavior without penalty identifies a social norm that the behavior will not be punished. The second theory that Professor Bandura advances to explain the influence of social modeling suggests that social models are even more important determinants of behavior. Professor Bandura advances a social cognitive explanation for the power of modeling.62 He argues that seeing a similar person engage in a behavior leads the observer to believe that they have the ability to engage in the same conduct. Social models show an observer what is possible, thereby changing what the observer is possible for someone like them to accomplish (even if the behavior is as simple as bashing a bobo doll). Observing similar others engage in conduct imbues people with a sense of personal (or self) "efficacy" concerning that conduct.63 Although self-efficacy is invaluable when it consists of a that one can pass an exam or drive a car, it also can produce a belief that one can rob a bank or blow up a building.61

58. See generally ALBERT BANDURA, AGGRESSION: A SOCIAL ANALYSIS (1973). 59. See id. 60. See id. at 72-85. 61. See id. at 85. 62. Id. 63. See BANDURA, supra note 58, at 399-400. 64. See id. at 431-33. 20001 THE LIMITS OF SOCIAL NORMS The social cognitive influence of social models and importance of self-efficacy indicates that groups have even greater power than law and social norms scholars have heretofore identified. The social information explanation implies that the power of social modeling is ephemeral and subject to manipulation. All that one would need to change an individual's behavior is to change the models. The self- efficacy explanation, however, indicates that social models are more insidious. Once someone gets a sense of personal efficacy as to a behavior, it is difficult to change their beliefs. Imbuing someone with the belief that they can successfully rob houses, for example, will not change just because a community starts to clean up trash and fix broken windows. Likewise, a person who quits selling drugs and obtains legal can inspire a more beneficial sense of efficacy in those who observe this behavior. Social models do not merely signal different norms, they affect how people view themselves and the possibilities in their lives. Thus, Professor Bandura's work indicates that social models have such lasting effects on behavior that simple manipulations of apparent norms might not really provide cheap, inexpensive solutions to antisocial conduct. Anticrime programs also have to worry about the kinds of models that a neighborhood provides. 3. Social Information Theories: Asch's Studies As the social cognitive theory of social modeling suggests, the influence of groups might be extremely difficult to manipulate. Fortunately, groups need not change people's perceptions or beliefs to change their behavior. The behavior of others can alter the meaning of social situations and dramatically affect their behavior, just as the law and social norms scholars predict. Consider a well-known study conducted by psychologist Solomon Asch in the early 1950s.65 In the basic design for his study, Asch asked subjects to participate in a group experiment on , ostensibly with seven other subjects.66 Only one of the subjects was a novice to the experiment, however, the other seven were actually Asch's confederates. The real subject was always seated with six subjects on his left and one on his right.67 The

65. See Solomon E. Asch, Opinions and Social Pressure, SCI. AM., Nov. 1955, at 31,31-35. 66. See id. at 32. 67. Although Asch only used male subjects, others replicated his results with female subjects. See Richard A. Crutchfield, Conformity and Character, 10 AM. PSYCHOL. 191, 196 (1955). CHICAGO-KENT LAW REVIEW [Vol. 74:1537 experimenter then informed the entire group that their task was to compare a target line to three lines of various lengths and determine which of the three choices was the same length as the target.68 The experimenter informed the subjects that he would show them a series of similar targets and choices, but the task was always to identify the line of the same length as the target. The experimenter always instructed the subjects to vocalize their answers in order such that the real subject responded only after six of the confederates had already responded. Asch's study began innocently enough with the first target; the answer seemed obvious, as the line lengths were quite disparate and all six subjects identified the line that seemed appropriate. As the experiment progressed, however, despite the fact that the right answer always seemed obvious, the confederates sometimes uniformly provided an obviously wrong answer with the same confidence and assurance as in previous rounds when they had provided the correct answer. The subjects in Asch's study found this behavior baffling. From their perspective, subjects in the same position as they were in clearly gave uniformly inaccurate responses, despite the ease of the task. In the face of this uniformity, subjects gave responses that conformed to the inaccurate response of the group on roughly one-third of the trials. 69 Most of the subjects (roughly eighty percent) gave responses that conformed to the group's inaccurate responses on at least one trial during the experiment.70 Asch's study could be said to demonstrate the overwhelming power of a group to create and enforce a social norm, thereby replicating Sherif's work. Even though the group consisted of complete strangers with no real power to sanction the subject, the subjects conformed to the group norm. Asch's work, however, illustrates a somewhat different point. The group did not influence the subjects' visual perception. Indeed, the very reason Asch used such obvious stimuli was to demonstrate that a group could change behavior without changing perception.71 Neither were Asch's subjects merely bowing to the pressure of a group.72 Rather, the group's

6& See Asch, supra note 65, at 32-33. 69. See id. 70. See id. at 33. 71. See id. at 32-33. 72. See Morton Deutsch & Harold B. Gerard, A Study of Normative and Informational Social Influences upon Individual Judgment, 51 J. ABNORMAL & SOC. PSYCHOL. 629, 632 (1955) 200 THE LIMITS OF SOCIAL NORMS inaccurate responses forced the subjects to reexamine their beliefs about the task. Subjects who conformed to the group's inaccurate responses did so because they ultimately determined that they might have misunderstood the task in some way.73 The group did not alter the subject's judgment of the stimulus materials, but it did alter the target of their judgment.14 Asch's study shows that groups can influence people's behavior without even changing their beliefs or norms.75 Two pieces of evidence confirm this interpretation of Asch's study. First, Asch conducted a version of the study in which one of the six confederates who responded before the subject always gave the objectively correct answers.76 In these instances, subjects rarely conformed to the majority of the group.7 7 In effect, the subjects disregarded the norm adopted by five of the six members of the group in favor of the minority. The single defector provided adequate assurance to the subject that they had correctly understood the task. Second, later replications of the Asch experiment showed that when subjects had an explanation for the group's anomalous responses subjects did not conform to the group norm.78 When the subjects came to believe that they had misunderstood the question, they conformed to the group; when they encountered either some support for their own views or another explanation for the group's behavior, the group's conduct had no effect on their own behavior.7 9 Asch's conformity experiments provide the best support for the law and social norms scholar's policy prescription on cleaning up neighborhoods as a way of reducing crime. Committing crime means something different in a clean neighborhood than in a dirty one. Even someone who does not subscribe to social norms against lawbreaking might reevaluate their behavior in a well-tended area. Improving a neighborhood's appearance need not affect a potential lawbreaker's judgment about the of committing crime in

(showing that even when subjects in Asch's responded anonymously, the group norm still influenced their responses). 73. See ROSS & NISBETT, supra note 52, at 34-35. 74. See id. 75. See generally Asch, supra note 65. 76. See id. at 34. 77. See id. 78. See Lee Ross et al., The of Attribution Processes in Conformity and : Re- Visiting the Asch Situation, 31 AM. PSYCHOL. 148, 150 (1977). The subjects were told that the confederates were operating under an unusual payment schedule that might have rewarded them for what seemed like inaccurate answers. 79. See Ross & NISBETr, supra note 52, at 34-35. CHICAGO-KENT LAW REVIEW [Vol. 74:1537 order to reduce crime; it need only change their judgment about the community's likely response to their behavior. 4. More Social Information Research: The Murder of Kitty Genovese Explained Psychologists have demonstrated the power of social influence to affect adherence to social norms in real-world settings as well. In particular, a series of studies by psychologists Bibb Latand and John Darley and others demonstrated that groups have a profound effect on the pro-social behavior of coming to the of a stranger.80 Their research was motivated by the shocking murder of Kitty Genovese in New York City. Ms. Genovese was murdered in an alley in New York while some thirty-six neighbors apparently watched from their apartment windows without so much as calling the police to help her. The case provoked and led many to question the morals of New Yorkers. The popular press argued that the incident demonstrated that New York City residents had developed a selfish social norm of refusing to aid their neighbors.81 To use the language of the law and social norms theorists, Latan6 and Darley challenged the theory that New Yorkers had abandoned the norm of coming to the aid of strangers in need, and argued instead for a social influence interpretation. Latan6 and Darley reasoned that when faced with a mildly ambiguous situation, such as unidentified screams from an alley outside one's window, people look to others to help them interpret the situation.82 Unfortunately, what they often see is other people looking to them for help interpreting the situation, rather than aiding the victim. This leads the bystanders to believe that the victim does not actually need assistance, or, if they do, the other bystanders must already have called for it. It is not that the bystanders subscribe to a social norm of not helping, it is that they unwittingly influence each other's interpretation of the situation. Latan6 and Darley and others conducted several empirical demonstrations of this phenomenon by simulating slightly ambiguous emergency situations. 83 In one such demonstration, they arranged for subjects to sit in a room completing a questionnaire while the room

80. See generally BIBB LATANfn & JOHN M. DARLEY, THE UNRESPONSIVE BYSTANDER: WHY DOESN'T HE HELP? (1970). 81. See id. at 1-3. 82. See generally Bibb Latand & John M. Darley, Group Inhibition of Bystander Intervention in Emergencies, 10 J. PERSONALITY & SOC. PSYCHOL. 215 (1968). 83. These are described in Latan6 & Darley, supra note 80. 20001 THE LIMITS OF SOCIAL NORMS slowly began filling with smoke.84 The subjects encountered this manufactured emergency either alone, with two confederates who sat passively as the room filled with smoke, or with two other naive subjects. When alone, seventy-five percent of the subjects called for help within the allotted for the experiment. When the subject was with two confederates who calmly continued to fill out the questionnaire as the smoke entered the room, only ten percent called for help. This result is striking, by the end of the experiment, the smoke had become so thick that it was difficult to see across the small room!85 Of the groups of three naive subjects, only thirty-eight percent reported a problem to anyone. 6 The presence of others inhibited a pro-social response. Latan6 and his colleagues replicated this result with more realistic stimuli. 87 In one study, they arranged for a female experimenter to lead male subjects into a room to be left either alone, with a male confederate, or with another naive male subject in a room connected to an adjacent room by an intercom system, ostensibly to complete a questionnaire. 88 After the experimenter left the room, she entered an adjacent room and, unknown to the subjects, turned on a tape recording of a staged accident. The recording contained sounds of the experimenter falling, complaining loudly about her ankle, and indicating that she was unable to move. The tape then continued with her moaning loudly for another minute. When alone, seventy-six percent of the subjects got help, or tried to determine what had happened, and the remaining twenty-four percent called out loudly to see if the experimenter needed assistance. When paired with a confederate who did not respond, only seven percent of the subjects reacted at all, and a majority of the pairs of naive subjects failed to respond. In later studies, Darley and his colleagues replicated these results with a staged epileptic fit.89 Finally, these researchers also showed that people who witness a crime (theft of a case of beer from

84. See generally Latan6 & Darley, supra note 82, at 215-18. 85. The result also raises some suspicion as to how ludicrous the behavior of the confederates must have seemed to the subjects, as the confederates were instructed not to acknowledge the presence of smoke in any way. See id. 86. See id. at 218. Because the base rate of reporting the problem was 75% per individual, 98% of the groups of three should have included at least one subject who would have reported the problem. 87. See Bibb Latand & Judith Rodin, A Lady in Distress: Inhibiting Effects of Friends and Strangers on Bystander Intervention, 5 J. EXPERIMENTAL PERSONALITY & SOC. PSYCHOL. 189, 189-93 (1969). 88. See id. at 192-93. 89. See LATANP & DARLEY, supra note 80, at 93-112. CHICAGO-KENT LAW REVIEW [Vol. 74:1537

a liquor store that occurred when a clerk was distracted) in the presence of others are far less likely to report the incident than when they are alone.90 In none of these studies can it really be said that the presence of others altered people's social norms about helping those in need. Neither was Kitty Genovese really the victim of a callous, urban social norm. Rather, the presence of a passive group affected the bystander's interpretation of Genovese's plight. The studies Latan6 and Darley conducted on bystander intervention reveal the importance of social signals in realistic settings. The results support the hypothesis that tidy neighborhoods can effectively signal whether a behavior will be tolerated, even if they do not alter beliefs about the appropriateness of committing antisocial acts. These experiments support the basic tenets of the law and social norms theorists. They show that norms develop quickly, and, if anything, have more power to influence behavior than the law and social norms scholarship indicates. This work suggests, in fact, that the power of norms derives not from implicit schedules of costs and benefits that groups bring, but from the power to influence perception and interpretation of a situation and even of oneself.

B. Limitations of Social Norms Theories The control that social norms exert on behavior is powerful, but phenomenally subtle. In some circumstances, people adhere so closely to a social norm that it seems unshakable. In others, however, social norms appear completely irrelevant. The latter point is best illustrated by social psychological studies in which the experimenters induce antisocial behavior that departs from significant social norms. Social psychologists have proven time and time again that, important though they may be, social norms do not completely dictate . Similarly, the prevalence of pro-social behavior depends upon seemingly unimportant social contexts.

1. Milgram's Obedience Experiments Psychologist Stanley Milgram conducted the most compelling demonstration that people will behave in ways contrary to their

90. See id. at 74-77. When alone, 65% of the witnesses reported the theft, while only in 56% of the two-person groups did even one of the witnesses report the theft. See id. at 76. 2000] THE LIMITS OF SOCIAL NORMS norms.9' Milgram's work demonstrated that ordinary individuals can be induced to brutalize others, against all norms of civilized conduct. Furthermore, Milgram induced antisocial conduct without changing the social meaning of the violent conduct; his subjects knew that what they were doing was wrong, and they did it anyway. Consequently, Milgram's study challenges social norms theories of any kind. Milgram conducted several versions of his study, but the most basic design went as follows. 92 The subjects consisted of ordinary residents of New Haven, Connecticut, who responded to a newspaper advertisement seeking subjects for a psychology experiment. Upon their arrival to Milgram's lab at Yale University, the subjects met two other individuals, one wearing a white lab coat introduced as the experimenter, and the other introduced as another subject. The other subject was, in fact, a confederate of the experimenter. The experimenter described the study as one that concerned the effects of aversive stimuli on learning and stated that he needed one of the two subjects to be the "teacher" and one to be the "learner." 93 The experimenter then conducted a rigged drawing that always assigned the role of teacher to the actual subject and the role of learner to the confederate. With the roles assigned, the experimenter explained the responsibilities of the teacher and learner. The learner proceeded to an adjacent room where the experimenter strapped him into a chair and connected him to wires capable of administering electric shocks. The teacher and experimenter would remain in a control room, and the learner would remain in an adjacent room. Contact was facilitated by an intercom system only. Ostensibly, the learner's task was to memorize a set of word pairs and to attempt to provide the correct responses. The experimenter explained to the teacher that his role would be to read the cue words, score the learner's response, and deliver an aversive stimulus to the learner for an inaccurate response. The aversive stimuli consisted of electric shocks administered through a control panel. The teacher was to change the setting of the shocks and press a button that triggered the shocks. The control panel made thirty shock settings available, labeled as ranging from 15 to 450 volts (in increments of 15 volts). Ranges of the shock settings were also labeled, from weak to strong as "slight shock, moderate shock, strong

91. See generally STANLEY MILGRAM, OBEDIENCE TO AUTHORITY (1974). 92. See id. at 13-26. 93. See id. at 18. CHICA GO-KENT LAW REVIEW [Vol. 74:1537 shock, very strong shock, intense shock, extreme intensity shock, danger: severe shock."94 The highest two levels were labeled "XXX." The experimenter explained to the teacher that he was to increase the shock intensity by one level with each incorrect response. 95 The experimenter also administered a forty-five volt shock to the teacher as a sample. The experiment began simply enough, with the learner scoring reasonably well, but with some inaccuracy. 96 This required the teacher to begin administering successively greater shocks to the learner. The learner also began to express escalating levels of discomfort with the shocks. At 75-105 volts, the learner grunted at each shock; at 120 volts, the learner verbally protested that the shocks were painful; at 150 volts, the learner shouted at the experimenter, complaining that he could not proceed; at 180 volts the learner shouted, " I can't stand the pain. '97 The progression up to 270 volts produced but and demands to be let loose. At 270 volts the learner began screaming in agony with each shock. At 300 volts he announced that he would refuse to cooperate with the experiment further, which he repeated at 315 volts. At 330 volts the learner screamed a last time and was not heard from thereafter. As the shocks began to increase, the subjects frequently protested to the experimenter. Milgram had devised a simple series of scripted responses for the experimenter to encourage the subjects to continue administering shocks.98 If the subject turned to the experimenter for advice or in protest, the experimenter was to give four responses in the following order: "please continue"; "the experiment requires that you continue"; "it is absolutely essential that you continue"; "you have no other choice, you must go on."99 If the subject asked about harm to the learner, the experimenter responded, "Although the shocks may be painful, there is no permanent tissue damage, so please go on."1° If the subject argued that the learner had refused to continue, the experimenter responded by saying, "Whether the learner likes it or not, you must go on until he has learned all the

94. Id. 95. See id. at 20-21. 96. See id. at 21-23. 97. Id. at 23. 98. See id. at 21-23. 99. Id. 100. Id. at 21. 2000] THE LIMITS OF SOCIAL NORMS word pairs correctly."'01 1 Other than these prompts, the experimenter made no demands of the subject. It is difficult to convey the drama of Milgram's experiment on paper; only Milgram's film of the subjects truly does the situation . The 1960s intercom system mutes the screams somewhat, but they are nonetheless chilling. The behavior of the subjects is testament enough. The subjects protest, wring their hands, laugh nervously, cry, and become angry at both the experimenter and the learner.1°2 Most of them also continued to administer painful electric shocks. In the version of the study described, no subject stopped administering shocks at less than 135 volts, and eighty percent administered shocks past 285 volts-well after the subject had asked to be let out of the experiment and begun screaming in pain.1°3 A surprising 62.5% administered all of the shocks all the way to 450 volts, several minutes after the learner had screamed in agony one last time and then ceased responding.1°4 Why were so many people willing to inflict brutal punishment on an innocent individual? It was not merely because Milgram had introduced a competing norm, as is often said of Milgram's study. There surely is a general norm of obeying authority in American society, but not to such a degree that people are willing to torture another because someone in a lab coat requests it. Authority matters, and Milgram found that people complied less when he conducted the experiment in a rundown section of New Haven, ostensibly on behalf of a fictitious organization named "Research Associates of Bridgeport."' 105 Even in this setting, however, forty-eight percent of the subjects completed the experiment by administering all of the available shocks. 1°6 Neither can one say that Milgram had set up his situation so as to change the social meaning of the subject's behavior. Torturing an innocent research subject is unambiguously immoral, and the subjects knew that their conduct was immoral. Their anxiety, protests, nervous laughter, and tears all attest to the fact that they believed that what they were doing was wrong. Many subjects repeatedly asserted that the study was immoral, even as they continued to deliver

101. Id. at 22. 102. See id. at 44-54. 103. See id. at 35. 104. See id. 105. See id. at 68. 106. See id. at 69. CHICAGO-KENT LAW REVIEW [Vol. 74:1537 shocks. 1°7 Also, in a version of the experiment in which the subjects were given the option of setting their own shock levels, they chose only the very lowest available, thereby revealing what ordinary people believe to be the appropriate behavior.tt s So, why did Milgram's subjects comply with the experimenter? Many facets of the experiment conspired to produce the result Milgram obtained, but there are two basic forces at work. First, the subjects were victims of a seductive social script that Milgram created to keep subjects in the experiment. When subjects requested that they be allowed to stop, Milgram's experimenter calmly insisted that the experiment required that they continue. Normally, a request to be let out of a situation would be honored, but Milgram had instructed the experimenter to respond passively and insist that the teacher continue administering electric shocks. When subjects offered to return the four dollars and fifty cents that they were paid to participate, the experimenter simply reminded them that they had earned the money just for showing up and that it was theirs to keep no matter what. When the subjects expressed concern for the well- being of the subject, the experimenter merely reassured the subject that the shocks could cause "no permanent tissue damage." 1°9 Milgram had closed the normal mechanisms of social interaction. To end the experiment, the subjects had to confront authority, but more importantly, they had to break out of the script that Milgram had arranged. They had to insist to the experimenter that he was wrong, that they did in fact have a choice, and that continuing was so inappropriate that their morality prevented them from continuing. Indeed, the subjects who refused to continue frequently apologized for being unable to go on, as if they felt they had wronged the experimenter in some way.110 Second, the subjects in Milgram's experiment had difficulty justifying their behavior to themselves. Milgram had made it easy for them to administer the first shock to the learner. The study seemed innocent enough, the task appeared easy, and the initial shocks were less painful than the sample shock that the subject had experienced. Once a subject administered the first shock, he or she was stuck. In order to stop administering shocks at any level, the subjects had to

107. See id. at 48-49. 108. See id. at 70-72. 109. Id. at 21-23. 110. See id. at 45-88. 20001 THE LIMITS OF SOCIAL NORMS justify their apparently immoral behavior up to that point. In fact, in all versions of Milgram's study, the subjects who stopped the experiment desisted at a point in which the condition of the learner had changed abruptly, such as when the learner screamed in agony, asserted that he did not want to be in the study anymore, or ceased responding altogether. There is no doubt that social norms played a role in Milgram's study. Milgram's script for the experimenter takes advantage of people's reliance on social norms. The typical requests to be let out of the experiment that the subjects made all consisted of appeals to social norms that the subjects clearly expected would be honored. When they were not, the subject was left with no conventional means to refuse the experimenter. Part of the reason the subjects remained in the experiment is that providing a reason for refusing a request from someone in the position of the experimenter is itself a social norm."' The other key reason for continued obedience is that the subjects wanted to believe that they were good, decent people, but if administering shocks was immoral, they had no way to justify the harm that they had already inflicted. Milgram's study could be said to have inverted the ordinary norms of social behavior to produce sadistic behavior in ordinary individuals. Milgram's study, however, reveals the power of situations to control behavior more than the power of social norms. The design of the experiment induced ordinary people to violate their social norms, and even break the law. Once the learner withdrew his to participate in the study, the subjects were committing a battery without experiencing any form of duress that the law would recognize as a sufficient defense. Continuing to administer the shocks also violates a profoundly important social norm of not harming an innocent person. The study suggests that even if social norms play an important role in social order, they can also produce undesirable conduct. Milgram's study indicates that espousing a pro-social set of norms may not be adequate to keep people from engaging in destructive conduct. Undesirable behaviors might often be produced by a complicated set of social circumstances that strongly resist change. If Milgram's study has general application, then crime might well be the product of people who face an unfortunate array of choices and opportunities in which law-abiding behavior does seem

111. See ELLEN J. LANGER, MINDFULNESS 14-15 (1989). CHICAGO-KENT LAW REVIEW [Vol. 74:1537 like a real option. Milgram's study suggests that antisocial conduct is the result of situations and not of an unfortunate set of norms. Thus, the results support the goals of a classic social liberal agenda of attacking poverty and developing opportunities for honest work, rather than working to change social norms.

2. The Ephemeral Influence of Norms: On the Road from Jerusalem to Jericho Subtle social factors can influence whether or not people conform to a pro-social norm as well. Innocuous background circumstances of a situation can dictate whether or not people will conform to a pro-social norm, thereby suggesting that social norms are less important determinants of behavior than more mundane aspects of social situations. A study conducted by psychologists Darley and Batson illustrates this point most clearly."2 The subjects of this study were undergraduate divinity students at Princeton University. Upon arrival to the experiment room, the subjects were told that the study concerned the ability of divinity students to think quickly, on their feet as it were, in preparation for a public speaking engagement. The experimenters told the subjects that they would have to walk over to another building and give a talk to a group of freshman divinity students. Half of the subjects were told to address employment opportunities for divinity students after graduation, and the others were told to discuss the parable of the good Samaritan. This manipulation was crossed with another variable that proved critical- the subjects were told either that they were already late for the talk and had to hurry, that they had just enough time to get to the talk, or that they had a few extra minutes. Darley and Batson's experiment truly begins during the subjects' walk over to the building to deliver their talk. All subjects passed a man who was slumped over against a wall, apparently in need of assistance. The man was, in , a confederate of the experimenters. As the subjects passed the confederate, he coughed twice and groaned. If the subjects asked him if he needed help, he said no, but it appeared otherwise. The subject of the sermon had no

112. See generally John M. Darley & C. Daniel Batson, "From Jerusalem to Jericho": A Study of Situational and Dispositional Variables in Helping Behavior, 27 J. PERSONALITY & SOC. PSYCHOL. 100 (1973). 20001 THE LIMITS OF SOCIAL NORMS effect on the rate of helping." 3 Whether the experimenter instructed the subjects to hurry or not, however, mattered a great deal. Subjects in a hurry were far less likely to stop and provide assistance than the other subjects. The results of the study are a stunning triumph of mundane features of a situation over social norms. The subjects were, after all, not a random sample of Princeton undergraduates who might lack a dedication to the social norm of helping those in need-they were divinity students. The beliefs that these students doubtless held dear, however, were easily manipulated from an instruction by an unknown experimenter to hurry. Furthermore, even making the parable of the good Samaritan salient had no real effect on the subjects relative to the instruction to hurry. A pro-social norm, it seems, has the most effect when acting on the norm is convenient. The results of Darley and Batson's study can be interpreted as consistent with social norms theories, and even with rational choice theories. Social norms theorists could argue that the experimenters invoked competing norms of fulfilling a commitment to the experimenter to not only give the talk, but to hurry over to the audience to do so. Similarly, rational choice theorists could argue that the experimenters increased the cost of tarrying for any reason. Both accounts are reasonable, but miss the important message of the study. The rational choice story clearly fails to give an indication of why the request to hurry increased the cost of helping to such an extent that the students were willing to abandon closely held beliefs. Similarly, arguing that the experimenters have introduced a competing social norm raises the question as to why one of the norms is so much more important than the other, and how anyone attempting to manipulate social norms to produce a certain behavior would be able to predict in advance which norm would have the most impact on behavior. One of the key lessons of Darley and Batson's study for social norms theories of crime control, and also the chief lesson of Milgram's study, is the notion of norms in tension. Given the results of these studies, social norms, even ostensibly pro-social ones, can be twisted to produce antisocial behavior.

113. Nor did a personality variable assessed by the experimenters, which was whether the subject was an intellectually-oriented divinity student or a spiritually-oriented student. See id. at 106. This latter variable, however, did influence the type of help that the subjects provided. See CHICAGO-KENT LAW REVIEW [Vol. 74:1537

III. LESSONS FOR LAW: CONCLUSIONS The decades of research in social psychology, as illustrated in the studies described in this Essay and in others, distill into three basic lessons. 114 First, people actively interpret the social circumstances in which they find themselves, making subjective reality as important a determinant of behavior as objective reality. In Asch's study, the objective reality of the line lengths did not dissuade subjects from conforming to the group norm as they struggled to make sense of this odd situation. Second, situations often overwhelm individual personalities and preferences and norms. Milgram showed that certain situations can overwhelm ordinary individuals, inducing them to engage in monstrous behaviors. Third, in most circumstances, multiple social forces push social behavior in opposite directions. Subjects in Darley and Batson's study believed deeply in the importance of helping others, but most ignored this belief in an effort to comply with a request to hurry. These basic lessons support the new law and social norms scholarship in many respects. The idea that group norms are powerful and important determinants of behavior pervades the social psychological literature. Sherif demonstrated that norms develop naturally, that people adhere to them even without apparent sanction, and that norms might even alter perception. Lewin showed how to use the influence of norms to induce pro-social behavior. Even the demonstrations of unusual behavior that Asch and Milgram produced take advantage of social norms. Clearly, even though they are sometimes inconsistent with each other, social norms are a powerful influence on social behavior. Hence, understanding social norms is a key to understanding social behavior. Similarly, the theory that others' conduct can alter the social meaning of behavior has ample support from social psychology. The basic concept of active, subjective interpretation of social settings is somewhat similar to the theories of social influence and social meaning that the law and social norms theorists have described. Asch's study shows that a group can induce people to reinterpret a social situation, just as a neighborhood can induce a potential lawbreaker to decide that committing a crime would be inadvisable there. Likewise, the passive responses of the bystanders in Latan6

114. The idea that these three basic lessons are the principle morals of the social psychological story comes directly from Ross & NISBETIr, supra note 52, at 8-17. 20001 THE LIMITS OF SOCIAL NORMS and Darley's studies show that people use the behavior of others to help them interpret social situations. In addition to the support for law and social norms theories, each of social psychology's three basic lessons suggests caveats to the new literature on social norms.

A. Social Influence Reexamined

Even though the law and social norms scholarship recognizes the importance of subjective social interpretations, its theories probably would not have predicted several of the results in the literature. For example, although the law and social norms theory would predict that a unanimous group will induce the naive subjects to reinterpret their situation in Asch's study, it probably would not have predicted that a single defector would drastically reduce the conformity effect. Game theory would doubtless predict that the 5-1 majority would win over many of the subjects,' 5 and yet most subjects easily resisted the group pressure in this condition. Similarly, the law and social norms theories would only have predicted part of the results in Darley's helping studies-those conditions in which the group consists entirely of confederates who refrained from helping. Where a group of subjects were all naive, the dominant social norm was doubtless to provide assistance, and yet fewer did so. Finally, although the results of Bandura's bobo doll studies are probably not a to the social norms scholars, the insidious effect that social modeling has on people's beliefs about themselves is not yet a part of the literature. Social models have a much more lasting impression than social norms scholarship recently admits. Although the basic lesson that groups influence behavior is clear in the social psychological literature, social influence is quite complicated. The key lesson of these observations for the law and social norms scholars is, ironically, not to underrate the power of social influence. People are dynamic and active interpreters of the social settings they encounter. Although people make many predictable errors of judgment in social settings,11 6 they actively search out meaning and make significant attributions about themselves that guide their subsequent behavior. Rather than search for the hidden schedule of costs and benefits that ostensibly give social norms their power to

115. See Picker, supra note 19 (developing models of social norms that assume increases in the strength of a majority will increase its influence). 116. See generally RICHARD NISBETr & LEE ROSS, HUMAN INTERFERENCE: SHORTCOMINGS OF SOCIAL JUDGMENT (1980) (documenting such errors). CHICAGO-KENT LAW REVIEW [Vol. 74:1537 influence behavior, the law and social norms scholars should attend to the influences that social psychology has documented. 117

B. Situationism The power of the situation to induce people to behave in ways that are contrary to their social norms does not yet seem to be a part of the law and social norms scholarship. In fact, this scholarship could accommodate the results of the studies by Milgram and Darley and Batson only by adopting a circular view of norms. That is, social norms scholars could argue that these two studies simply introduced more powerful social norms that overcame norms of decency. This view would not tell a policymaker much about when one social norm might triumph over another, at least not until after the fact. The better view of the research conducted by Milgram and Darley and Batson is that people tend to underestimate the coercive power of situations to trump people's norms and preferences. The studies are interesting sides of the same coin. Milgram demonstrated that people are capable of destructive, even deadly, conduct under circumstances that seem benign. Milgram's subjects did not behave this way because they are people who espouse a destructive or violent set of norms; they behaved this way because of the situation Milgram created. Similarly, Darley and Batson's subjects did not help solely because they were people who espouse altruistic norms, they helped because they espoused such norms and had some extra time. Situationism is particularly important for the social policy implications of the social norms theories. Law and social norms theorists argue that a neighborhood can cheaply send signals to potential lawbreakers that will redefine their conduct by showing that the neighborhood will not tolerate crime. Milgram's data suggest, however, that crime is not a product of fear of retribution or an undesirable character, but rather a product of a somewhat arbitrary set of social pressures. Milgram managed to turn ordinary people into criminals in under twenty minutes. An environment that presents opportunities to commit crime and presents few obvious alternatives to a comfortable life surely has an even greater affect on decisions to engage in antisocial conduct. Tidying up neighborhoods

117. But see Richard A. Posner, Social Norms, Social Meaning, and Economic Analysis of Law: A Comment, 27 J. LEGAL STUD. 553, 563 (1998). "[T]he subject of social norms can be analyzed fruitfully in terms of economic theory conceived as the theory of rational individual choice rather than as the study of conventional economic markets." Id. 2000] THE LIMITS OF SOCIAL NORMS might deter some crime, but it is unlikely to channel people's behavior into socially acceptable paths.

C. Social Norms in Tension The concept of social forces in tension suggests a pragmatic, but optimistic, point about social norms. Social norms are truly ubiquitous, and often in conflict. Therefore, there are two ways to increase the rate of a desirable social behavior: either augment the social forces that increase its prevalence or dampen those social forces that reduce its prevalence. In other words, sometimes the question is not what encourages people to do something, but rather what discourages them from doing it. Often, social reforms emphasize the former issue to the neglect of the latter. The law and social norms scholarship has done the same, to some extent. It has claimed that slovenly neighborhoods make crime seem more attractive than it really is, or should be, and therefore, a slovenly neighborhood encourages people to commit crime instead of find honest work. It may be that the people do not find honest work are more important determinants of criminal behavior than the apparent attractiveness of crime.

D. Conclusion

Social psychology has a long and rich history of research on social norms and how they function. Rather than incorporate this work, the new scholarship on law and social norms has relied instead on economics and game theory as its theoretical underpinnings. As valuable as the insights from economics and game theory are, they paint a somewhat impoverished portrait of humans as social animals. It is certainly true that one of the reasons people conform to group norms is the rewards of belonging that come with a group, be these rewards financial remuneration or an enhanced sense of self-worth. People are such subtle and active processors of their social worlds, however, that the analysis must go further to reap any analytic fruit and have any predictive power. The basic lessons of social psychology would do much to provide the predictive power that law and social norms scholarship will otherwise lack.